If your AWS Elemental MediaLive channel is sending video to YouTube but you cannot hear sound, trace the audio path from the source to YouTube before changing settings. The symptom alone does not identify the cause: audio may be absent at the source, not selected by the input, unmapped in the output, rejected by the ingest workflow, or simply muted on playback.
Work through those boundaries in order and note where audio is present or disappears. Keep the checks conditional on your input type and output protocol; an HLS example is not a universal recipe for an RTMP channel.
1. Confirm that the source contains audio
Start before MediaLive. Listen to the source itself or inspect a recording or local monitor from the device that supplies it. If the source is a video file, play the actual file and check that the intended section has audible sound. If it is a camera, mixer, or another live feed, check the relevant monitor output and confirm that the audio feed is connected and active. A picture at the source does not prove that an audio signal is present.
If there is no audio at this point, changing a YouTube setting or MediaLive output will not restore it. Check the source device, its audio connection, and any upstream mixer or routing that feeds it. For a playlist channel assembled from recordings, inspect the files themselves rather than assuming every item has the same tracks; guidance on preparing lecture recordings for a continuous live playlist is relevant when the source is a collection of recordings.
For an AWS Elemental Link source, AWS documents signal lock and the EncoderRunning metric as checks. Its MediaLive User Guide says a value of zero indicates the encoder is not processing the input. If that is what you see, verify that the device is receiving a valid signal rather than trying to repair the symptom in YouTube. Consult the AWS Elemental MediaLive User Guide for current details and confirm which source type your channel uses.
Write down what you have established: whether the source is audible, which audio track or language is intended, and whether the source is a file, a device, or an Elemental Link. Those details determine what to inspect next. If several inputs are used, repeat the source check for each one that can be active; one working input does not establish that every alternate input carries audio.
2. Check the input signal and encoder status
Once you have verified audible source content, check what reaches the MediaLive input. The goal is to establish whether the expected signal arrives at the input, not to infer audio from the fact that video is present. Use the monitoring available for your source and input type, and compare the current input with the source you just checked. If the input has no usable signal, the fault is still upstream of output encoding.
For Elemental Link, check signal lock and EncoderRunning as appropriate. AWS’s documented interpretation of a zero EncoderRunning value is that the encoder is not processing the input. That points you towards checking the valid signal passed to the device. It does not, by itself, say what is wrong with another input type or prove that a particular cable or setting is at fault.
For other inputs, inspect the channel’s attached input and the status or monitoring that is available for that source. The exact indicators depend on the input configuration, so do not look for a metric from the Link example on a different type of input. If video is reaching the channel but audio is not, continue to the audio selector checks; the picture alone cannot tell you whether MediaLive has access to the intended audio asset.
Keep a simple record of the input attachment, source type, and whether the input is currently active. If the channel switches between sources, check the one that is actually on air when the symptom occurs. This helps avoid changing a selector on an inactive input and then mistaking a later source change for a fix. A 24/7 playlist stream setup has the same practical concern: know which source is feeding the broadcast before diagnosing its output.
3. Inspect MediaLive input audio selection
MediaLive audio selectors tell the channel which audio asset to extract from an input. Check the input attachment used by the channel and identify the selector configured for the intended audio. Then verify that the selector corresponds to the actual asset available from that input: depending on input type, that can mean checking the selected track, PID, or language.
Selector requirements vary by input type. AWS describes RTMP as carrying a single audio asset, while HLS and some multi-track input types may require the selector to identify a PID or language. Do not copy a selector value from a different input or assume that a selector which works for one attachment will identify the same track on another. The MediaLive User Guide explains selectors and input-specific requirements; use it alongside the configuration actually attached to your channel.
If the channel has multiple inputs and sound is missing only when a particular one is active, compare that input’s selector with the others. AWS notes that a multiple-input channel must extract the same audio languages from each input and use identical selector names across inputs. That makes both the selected language and the selector naming worth checking in a failover or switching setup. It does not mean every channel should be configured with multiple selectors or a particular language.
If a selector is missing or points at the wrong asset, correct only the relevant input selection and then observe the result before editing the output. Record the original selection so you can restore it if the change does not help. If the input is a single-audio RTMP source, a multi-track PID assumption is unlikely to be the useful next check; follow the requirements for that input type instead.
4. Verify output audio mapping and encoding
After confirming the input selection, inspect the output that sends the stream to YouTube. Confirm that an audio encode exists in the output and that it references the intended input audio selector. A channel can have an audio track available on its input and still fail to send it if the output does not include or map an audio encode.
Then check the output group and audio format against the ingest protocol configured for your YouTube stream. AWS’s documented MediaLive-to-YouTube HLS example uses an HLS output group, selects Audio 1 in the stream settings, and uses AAC. AWS published that example in 2021, so treat it as a specific HLS configuration illustration rather than an instruction to apply unchanged to every current channel. See the AWS Media Blog HLS example and verify current console labels and supported choices in the documentation for your configuration.
That distinction matters if your channel uses RTMP. YouTube’s live encoder guidance uses a stream URL and key, but the correct MediaLive output path depends on the workflow you configured in YouTube Live Control Room. Do not take the HLS mapping example and apply it to an RTMP output merely because both send video to YouTube. Check the protocol and destination as a pair, then check that the audio encode is attached to the matching output.
If you intend to send ordinary mono or stereo audio, do not change settings for surround sound without a reason. YouTube’s guidance for 5.1 live audio is a separate case: it specifies compatible content and an encoder, and lists protocol-specific format requirements. For 5.1, YouTube lists AAC for RTMP and AAC, AC-3, or E-AC-3 for HLS, with 48 kHz and 384 Kbps. These figures apply to the stated 5.1 guidance, not as general requirements for every stereo stream. Check the current YouTube 5.1 live audio guidance if 5.1 is what you are sending.
5. Check YouTube ingest and monitoring
Confirm that MediaLive is sending to the intended YouTube workflow and destination. Compare the output group with the ingest protocol configured in Live Control Room, and check that the intended stream URL or destination and stream key are being used. YouTube’s live encoder setup guidance describes configuring the encoder with YouTube’s stream URL and key. A mismatched destination can make observations from one stream irrelevant to another, so confirm which broadcast you are actually monitoring.
Now compare the sound at the source with the live preview in YouTube Live Control Room. If the preview has no audio, continue upstream through the output mapping, selected audio, and input. If the preview has audio but viewers or your local player hear silence, check the playback device, player volume, mute state, and playback path before changing the MediaLive channel. This comparison narrows the search; it does not establish a particular cause without further checks.
Be careful to distinguish a silent preview from a silent playback device. Check the player’s mute control and volume, the device’s output selection, and whether another playback path can reproduce the stream. If possible, compare the same live output in the Control Room and on the affected viewing device. If those disagree, changing the MediaLive selector may introduce a new fault while leaving the playback issue untouched.
If you use a continuous channel, keep the stream configuration and source notes together so an operator on another shift can tell which input, output group, and destination were tested. The same discipline is useful when setting up a continuous devotional stream: a clear handover makes it easier to separate a change in programme source from a change in the live path.
6. Change one thing at a time
Once you have located the boundary where audio disappears, make the smallest applicable change there. For example, if the source is audible and the input selector is clearly aimed at the wrong asset, correct that selector and observe the output. If the selector is right but the output has no audio encode or maps a different selector, address the mapping instead. Avoid changing the selector, codec, output group, and YouTube destination together; if audio returns, you will not know which change mattered, and if it does not, you will have more settings to unwind.
Before editing, capture the relevant current configuration: input attachment and type, selector name and selection, output group, audio encode and mapping, and destination or protocol. You do not need to record every channel setting. Capture the parts of the path involved in the test so you can compare the state before and after. If you are working with a team, note who made the change and when in your own operational log rather than relying on memory.
After each change, check the same observation points in the same order: source, input, output mapping, YouTube preview, then playback device. Give the stream the time your normal workflow requires to reflect a change; do not treat a brief transition or a single local playback check as proof of a stable result. If the expected point still has no sound, restore the prior setting if appropriate and move to the next unresolved boundary.
Do not treat a successful test on one input as proof that all inputs are correct. For a channel with alternate sources, test each relevant input and the transitions between them, especially if selector naming or languages differ. Nor does a sound-bearing preview guarantee that every viewer’s device or player path is audible. The purpose of this sequence is to isolate the fault and keep changes attributable, not to promise that a particular configuration will work in every channel.
If the cause turns out to be the burden of keeping a file-based broadcast running while your computer is off, that is a separate operational problem from diagnosing a MediaLive audio path. StreamNeo can remove the need to keep your own computer running for a file-based YouTube stream; it does not replace checking source audio or prove that a MediaLive configuration is correct.
Before committing, compare the operating options on the pricing page. When the file and channel are ready, start free — 24-hour trial, no card.
FAQ
Why is video reaching YouTube when audio is not?
Video and audio take related but distinct paths through the source, MediaLive input selection, output encoding, and YouTube ingest. Video arriving does not prove that the intended audio asset was selected or mapped to an output. Check each boundary in order rather than assuming the failure is at YouTube.
Should I use AAC for every MediaLive YouTube stream?
No single setting applies to every channel. AWS’s published HLS example uses AAC for its MediaLive-to-YouTube HLS workflow, while the appropriate output depends on the ingest protocol and audio format you are using. Check the current guidance for that workflow before changing the encode.
What should I check if only one input has no sound?
Compare that input’s source audio, input attachment, selector, and output mapping with an input that works. For multiple-input channels, AWS calls for the same audio languages to be extracted and identical selector names across inputs. Confirm the actual channel configuration before changing it.
When should I change YouTube settings?
Check the YouTube destination, protocol, and stream key when confirming that the intended MediaLive output is reaching the intended live stream. If the Control Room preview has sound but a viewer hears silence, check playback and device mute first. Change YouTube or MediaLive settings only when the observations point to that part of the path.